Response to the following problem:
One of the extensions to the JPEG standard allows for the use of multiple quantization matrices. Investigate the issues involved in designing a set of quantization matrices. Should the quantization matrices be similar or dissimilar?
How would you measure their similarity? Given a particular block, do you need to quantize it with each quantization matrix to select the best? Or is there a computationally more efficient approach? Describe your findings in a report.